B763 / B744, Amsterdam Netherlands, 1998

B763 / B744, Amsterdam Netherlands, 1998

Summary

A Boeing 767-300 departing from runway 24 at Amsterdam made a successful daylight rejected take off upon seeing a Boeing 747-400 under tow crossing the runway ahead. It was found that the crossing clearance had been given by the same trainee controller who had then cleared the 767 for take off after assuming that the towing traffic had cleared based on an unverified assumption based upon incorrect information which had been received earlier from an Assistant Controller. The conflict occurred with LVP in force and with visual surveillance of the runway from the TWR precluded by low cloud.

Event Details
When
10/12/1998
Event Type
HF, RI
Day/Night
Day
Flight Conditions
On Ground - Normal Visibility
Flight Details
Operator
Type of Flight
Public Transport (Passenger)
Take-off Commenced
Yes
Flight Airborne
No
Flight Completed
No
Phase of Flight
Take Off
Flight Details
Operator
Type of Flight
Out of Service
Take-off Commenced
No
Flight Airborne
No
Flight Completed
No
Phase of Flight
Pushback/towing
Location
Location - Airport
Airport
General
Tag(s)
Aircraft-aircraft near miss, Inadequate ATC Procedures
HF
Tag(s)
ATC clearance error, ATC Unit Co-ordination, Ineffective Monitoring, Procedural non compliance, Spatial Disorientation
RI
Tag(s)
ATC error, Incursion pre Take off, Runway Crossing, Towed aircraft involved, Near Miss, Phraseology
Outcome
Damage or injury
Yes
Non-aircraft damage
Yes
Non-occupant Casualties
No
Off Airport Landing
Yes
Ditching
Yes
Causal Factor Group(s)
Group(s)
Air Traffic Management
Safety Recommendation(s)
Group(s)
Air Traffic Management
Airport Management
Investigation Type
Type
Independent

Description

On 10 December 1998, Boeing 767-300 being operated by Delta Air Lines on a scheduled passenger flight from Amsterdam to Atlanta GA rejected its take off from runway 24 in misty daylight conditions after its crew saw a Boeing 747-400 being towed across the runway ahead of them. The aircraft was brought to a stop before reaching the 747.

Investigation

An Investigation was commenced by the Accident and Investigation Bureau of the Netherlands Aviation Safety Board. It was established that both aircraft had been issued with clearances which corresponded with their actions. The weather was overcast and surface visibility was being reported as 1700 metres in Mist with runway 24 RVR ranging between 1600 and 1800 metres. It was not possible to monitor ground movements from the VCR and Low Visibility Procedures (LVP) were in force.

About 15 minutes prior to the conflict event, the runway configuration had routinely changed from two landing runways (19R and 01R) and one departure runway (09) to two take off runways (09 and 24) and one landing runway (19R). Positions active in respect of runway 24 were one controller who was also responsible for runway 09, a trainee supervised by an OJTI who was also designated as TWR supervisor, a Ground Controller responsible for ground traffic on the southern side of the airport including the area around runway 24 and an Assistant Controller who was responsible for communications with all vehicles on the whole airport manoeuvring area in Dutch.

The Assistant Controller received a request for a Boeing 747 under tow to cross runway 24 and the Runway Controller instructed that it should be held since an aircraft was about to roll. He was told by the Assistant Controller that the tow was waiting to cross from west to east at exit 2. Once the departing aircraft had become airborne from 24, The Runway Controller advised the Assistant Controller that the tow could cross. The tow leader then requested that the illuminated stop bar be extinguished but when this was attempted by the Runway Controller, it was realised that the lighting panel was still set up for the previous runway configuration. Once this had been rectified, the Runway Controller selected the west side stop bar to unlit but was surprised to hear the tow leader say that it hadn’t gone out. With the consent of his OJTI, the Runway Controller then selected both east and west side stop bars at exit 2 to unlit.

The Runway Controller then began to consider whether it would be possible to get the Delta 767 lined up at the beginning of runway 24 airborne in a rapidly reducing ‘window’ which had appeared because of a gap in landing traffic on runway 19R upon which runway 24 departures were conditional in the prevailing runway configuration. Having seen a ground radar target on the east side of exit 2 which he assumed was the tow traffic now clear and noticing that both exit 2 stop bars were now showing lit again on the lighting panel, he looked across to the Assistant Controller for confirmation but was unable to get her attention straight away. Realising that the take off ‘window’ for the Delta departure was about to expire and assuming that the crossing tow traffic was indeed clear, he issued take off clearance to the 767.

Twenty seconds later, the 767 reported rejecting the take off and it could be seen from the ground radar that it had slowed to an almost complete stop somewhere between exit 4 and exit 3. Only when the 767 announced that the KLM aircraft ahead was being towed did the Controller realise what had caused the aborted rejected take-off and he then “noticed for the first time that the car and tow were crossing in the direction opposite to what he had been expecting. The OJTI had not heard the take-off clearance given to the 767 “possibly as a result of his involvement in the discussion about the allocation of the stop bar control panels. He was therefore completely surprised by the situation. He could see the 767 stopping and also tow traffic crossing the runway from east to west now well across the runway but not yet clear.

The Assistant Controller had been looking at the ground radar display at the position to her right in order to see where the crossing traffic was but had been unsuccessful. The Assistant position has no radar display and the available displays nearby are set by the Runway Controllers to their preference which is not necessarily useful to the Assistant Controller responsible for vehicle communications.

It was determined that the conflict had occurred because of the wrong assumption about the actual position and direction of crossing of the towing traffic. The tow traffic had not advised its actual position or destination, the call made only – in line with prevailing procedures, a request to cross runway 24 at exit 2. It was observed that the Assistant Controller had not asked for clarification and “assumed for no specific reason” that (the tow traffic) was waiting at exit 2W of runway 24 with its destination the south apron and then passed that information to the 24 Runway Controller.

Route of DAL039 and KLM B747, reproduced from the Official Report

The Conclusion of the Investigation was as follows:

Only due to a reasonable actual visibility along the take off direction of the runway and quick and proficient action by the 767 flight crew in rejecting their take-off was a catastrophic accident was avoided.

It was also noted that:

  • the design and position of the control panels for stop bars and traffic lights are not unambiguous and therefore prone to human error
  • The non-use of checklists during the change-over from inbound- to outbound mode resulted in an initially wrong set-up for the stop bar control panel in relation to the Controllers’ duties. This reinforced their doubt about the correct functioning of the system and interfered with their recognition of the wring assumption about the position and movement of the tow;
  • OJTI (and TWR Supervisor) failed to adequately supervise Tower operations in general and did not make a timely intervention to prevent the conflict
  • The personnel on duty in the Tower were not working as a team.

The following Causal Factors for the conflict were identified:

  • Low visibility weather conditions which prevented Air Traffic Control from visually identifying vehicles on the ground
  • Inadequate information during the radio communications between the tow leader and the Tower
  • Misinterpretation of the position and direction of movement of the tow
  • Take-off clearance given without positive confirmation that the runway was unobstructed
  • Insufficient teamwork and supervision

Nine Safety Recommendations were made as follows:

  • that the Airport Authority should ensure that technical facilities with regard to the protection of runway exits are identical to allow standard procedures for all runway crossings, especially with regard to exit 2 of runway 06/24. In the meantime movements to/from the South apron other than by taxiing aircraft should not be allowed during low visibility weather conditions. In this connection it is recommended to follow up ICAO Annex 14 Standards as soon as possible.
  • that the ANSP and the Airport Authority should ensure that refresher training of procedures and radio communication should be provided to ATC Tower staff and Airport Authority manoeuvring area workers
  • that the ANSP should ensure that the Tower Supervisor does not have additional duties
  • that the ANSP should ensure that checklists are used when changing runway configuration
  • that the ANSP and the Airport Authority should ensure that the control panels for stop bars and traffic lights are redesigned and integrated geographically to avoid any ambiguity
  • that the ANSP should add a recording capability to the existing ground radar
  • that the ANSP should ensure that Assistant Controller positions are equipped with a multi-mode display screen
  • that the ANSP should ensure that Team Resource Management training is implemented for Air Traffic Control staff
  • that the ANSP and the Airport Authority should ensure that the effectiveness of co-ordination and communication procedures between the two organisations are re-evaluated

The Final Report of the Investigation was published by the Dutch Transport Safety Board (which had become responsible for the Investigation whilst it was in progress after a change in the organisation of safety investigations) in January 2001. The substantive Report is in English but also includes an abbreviated Dutch language version.

Further Reading/Related Articles

SKYbrary Partners:

Safety knowledge contributed by: